[Talk-us] Bridges, nodes, and routing engines (Navit, Gosmore, etc

[Alan Millar]

> Now the question is: should I remove the shared nodes (or detach them
> in JOSM),

Yes, delete or detach.  For editing simplicity, I detach them and at least
move one a few feet away.  In my region, the overpasses and bridges are
usually misplaced or distorted, and need to be moved (or completely
rebuilt) anyways.

> fixed?  My gut feeling is that the routing engines should be smarter
> (basically, don't allow a route to change layers except at the ends of
> OSM ways - in other words, you can't change from layer=0 to layer=1 by
> turning off a way - in other words, any sane turn off a way not at its
> end can't take you across layers,** and you can only "descend" or
> "ascend" layers at the ends of OSM ways) but maybe this is a cleanup
> that needs to be done in the data instead.

That's a lot of assumptions in a lot of routing engines, for roads that
just aren't connected at that node.  I think we want the most useful data
that can be readily used in as many different routing engines as possible,
without a lot of obtuse rules and transformations.  IMHO, If the roads
don't really connect, don't share a node.

> I've been reluctant to remove the nodes since we may need them to
> match up other TIGER data (for example, the addressing data really
> needs the tilds off the nodes to be able to figure out how to
> associate addresses with the existing imported linework- the way tlids
> alone are way too imprecise for this in many cases).

Right now, I don't think keeping the nodes will make a big difference in
retrofitting the TIGER addressing data.  The TLID really applies to a
Tiger line (way) section, not node, so the intersection nodes already have
multiple TLIDs on them (some nodes have a dozen or more where long ways
connect).  And once the data got into OSM, ways started being split up and
rejoined, without TLIDs being edited to match, so it is more of a hint
than concrete identifier.

What might really help for accurate addressing import is the Tiger TZID (a
unique node identifier) on each node, but we didn't get that on our bulk
load.  I don't know if it is a more recent addition to the Tiger data or
if we just didn't foresee it.  Then again, it may not survive multiple
edits any better.

I would guess that in most of these situations, the bridge, overpass, or
tunnel doesn't have any useful addressing that we need to salvage, so
having the tiger-related node on the street level is probably fine.

- Alan